Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

A graph attention network-based multi-agent reinforcement learning framework for robust detection of smart contract vulnerabilities.

Scientific reports·2025
Same author

LAVRF: Sign language recognition via Lightweight Attentive VGG16 with Random Forest.

PloS one·2024
Same author

EnViTSA: Ensemble of Vision Transformer with SpecAugment for Acoustic Event Classification.

Sensors (Basel, Switzerland)·2023
Same author

Plant-CNN-ViT: Plant Classification with Ensemble of Convolutional Neural Networks and Vision Transformer.

Plants (Basel, Switzerland)·2023
Same author

Recent Advances in Traffic Sign Recognition: Approaches and Datasets.

Sensors (Basel, Switzerland)·2023
Same author

HGR-ViT: Hand Gesture Recognition with Vision Transformer.

Sensors (Basel, Switzerland)·2023

相关实验视频

Updated: Jul 27, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

3.9K

用1D卷积神经网络对人类行动识别进行微调时间密集采样.

Kian Ming Lim1, Chin Poo Lee1, Kok Seang Tan1

  • 1Faculty of Information Science and Technology, Multimedia University, Melaka 75450, Malaysia.

Sensors (Basel, Switzerland)
|June 10, 2023
PubMed
概括
此摘要是机器生成的。

这项研究引入了一种新方法,即使用1D卷积神经网络 (FTDS-1DConvNet) 微调时间密集采样,用于识别人类行为. FTDS-1DConvNet显著提高了对基准数据集的分类准确性.

关键词:
一维卷积神经网络 (1D ConvNet)1D-CNN 1D-CNN 是一个数字.开始 - 恢复网-V2人类行动承认承认时间密集采样.

更多相关视频

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

586
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.5K

相关实验视频

Last Updated: Jul 27, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

3.9K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

586
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.5K

科学领域:

  • 计算机视觉 计算机视觉
  • 机器学习 机器学习
  • 人工智能的人工智能

背景情况:

  • 人类动作识别对于许多应用程序至关重要,但由于视觉变化而面临挑战.
  • 先进的表示学习提高了性能,但强大的动作识别仍然很困难.

研究的目的:

  • 为了应对视觉变化引起的人类行动识别的挑战.
  • 提出一种新的方法,FTDS-1DConvNet,用于增强特征提取和分类.

主要方法:

  • 时间分段将视频分成部分.
  • 时间密集采样和微调的Inception-ResNet-V2用于特征提取.
  • 1D卷积神经网络 (1DConvNet) 用于表示学习和分类.

主要成果:

  • 在UCF101数据集上实现了88.43%的分类准确度.
  • 在HMDB51数据集上实现了56.23%的分类准确度.
  • 在两个数据集上都超过了现有的最先进的方法.

结论:

  • 拟议的FTDS-1DConvNet方法有效地捕捉了人类行动识别的突出特征.
  • 与当前最先进的技术相比,FTDS-1DConvNet表现出卓越的性能.